Protease inhibitors represent one of the most successful and important components of current antiretroviral therapy. They are the only target for which tight-binding transition state analogs have been developed. High level resistance requires the evolution of numerous mutations, indicating a high genetic bar. For these reasons, protease inhibitors will continue to be a critical part of potent therapy against HIV-1. Understanding drug resistance to protease inhibitors is key to using the available inhibitors and designing new inhibitors with broader spectrum. Our knowledge of the effect of individual mutations that contribute to resistance is fragmentary, and while major themes of mechanisms have been revealed, the placement of these mechanisms in the complex setting of resistance evolution and the interaction between different resistance mutations is incompletely understood. Database searches (Project 1 - Shafer) provide a powerful tool for identifying positive and negative linkages between mutations that occur in the presence and in the absence of the selection for resistance. We will use such database searches as a starting point for identifying single mutations and linked mutations for characterization. We will apply a series of highly sensitive assays to reveal changes in virus resistance and fitness that are associated with the small effects of single mutations. Features of linkage and antagonism with be explored with these same assays in an attempt to understand the mechanistic pathway the protease follows in the evolution of high level resistance. Understanding the functional consequences of these interactions will provide a strong basis for interpreting models incorporating linked mutations (Project 4 - Gibson) and, in some cases, structures (Project 3 - Schiffer). An extreme form of fitness penalty for resistance mutations appears to be the development of hypersusceptibility to other selected protease inhibitors. We will determine if this effect is due to enhance inhibitor binding or to reduced enzyme activity. We will explore two models of fitness loss, poor enzyme activity leading to under-processed NC-pl and poor incorporation of RT. We will also examine the rate requirements for processing at the NC-pl site as a determinant of infectivity. Finally, we will develop SIVsm as a resistance/fitness model to prepare for future studies that will allow a formal test of the contribution of attenuation replication capacity to reduced pathogenicity. As a service to the drug development effort (Projects 4 and 5), we will provide virology support in characterizing new inhibitor leads.